China

Highlights February’s credit release earlier this week confirmed that credit growth is not yet on a “blowout” trajectory. If maintained, the recent pace of credit expansion implies a moderate credit cycle, not a large acceleration like what occurred in 2015/2016. We agree that a trade deal between China and the U.S. is likely to occur, but a sustained, cyclical (i.e. 6-12 month) rise in Chinese relative equity performance requires stability in the outlook for earnings, which have not yet reflected the ongoing economic slowdown. A confirmed meeting date between Presidents Trump & Xi coupled with more evidence that a moderate credit expansion is underway would likely lead us to upgrade our cyclical stance towards Chinese investable stocks (to overweight). Feature Tables 1 and 2 on pages 2 and 3 highlight key developments in China’s economy and its financial markets over the past month. On the growth front, data releases later this week will provide a crucial read on the pace of the slowdown in coincident economic activity. The ongoing weakness in trade and producer prices suggests that activity has continued to decelerate as the previously beneficial trade frontrunning effect washes out of the data. While we agree that January’s gargantuan credit number means that growth will bottom at some point this year, the February data released earlier this week highlights that credit growth is not yet on a “blowout” trajectory. If maintained, the recent pace of credit expansion implies a moderate credit cycle, not a large acceleration like what occurred in 2015/2016. Table 1China Macro Data Summary Table 2China Financial Market Performance Summary From an investment strategy perspective, we recommended in our February 27 Weekly Report that investors place Chinese investable stocks on upgrade watch, but that an immediate shift to a cyclical overweight was not yet warranted. The recent outperformance of investable stocks vs. the global benchmark largely reflects global investor expectations of a trade deal between China and the U.S. in the very near future, which we agree is likely to occur. But we have underscored that a sustained, cyclical (i.e. 6-12 month) rise in Chinese relative equity performance requires stability in the outlook for earnings, which have not yet reflected the slowdown that is underway. Barring a substantial trade-deal-driven rise in the RMB (which would dampen profits further and raise the bar for credit), a confirmed meeting date between Presidents Trump & Xi coupled with further evidence that a moderate credit expansion is underway would likely lead us to upgrade our cyclical stance towards Chinese investable stocks (to overweight). In reference to Tables 1 and 2, we provide several detailed observations concerning developments in China’s macro and financial market data below: The January and February data for several measures of coincident activity, including both measures of the Li Keqiang index (LKI) that we track, are set to be updated tomorrow. However, a number of data series that have been released over the past two months point to a continued deceleration: growth in rail cargo volume ticked down in January, producer prices are on the cusp of deflation, and nominal import and export growth decelerated again in February (measured either in US$ or RMB terms). The four components of our LKI leading indicator available for February have all sequentially declined, including the growth in adjusted TSF and adjusted TSF as a share of GDP. Credit had surged in January, but ticked down in February. Chart 1 illustrates the likely path of adjusted TSF as a share of GDP if the average pace of credit growth over the past three months is sustained. The chart implies that credit will have durably bottomed, but that the pace of advance will be weaker than that experienced in past cycles. Chart 1The Recent Pace Of Growth Implies A Moderate Credit Cycle​​​​​​​ January and February data for residential floor space started and sold will also be updated tomorrow, and it will be important to see whether the gap that has emerged between construction and sales has persisted. Floor space sold has reliably led starts since 2010, and we recently highlighted that the PBOC pledged supplementary lending program has led sales since 2015. The pace of PSL decelerated further in February, suggesting that the outlook for sales (which are already in negative YoY territory) is deteriorating. Based on the leading relationships that we have identified, residential construction volume is unsustainably strong. The seemingly inconsistent messages between the NBS and Caixin manufacturing PMIs in February (down and up, respectively) may in fact reflect the PBOC’s focus on easing financial conditions for small businesses. While the NBS PMI includes a much broader sample of firms than the Caixin PMI, the latter focuses heavily on private sector SMEs. Given this, February’s data may suggest that the export outlook is improving, but we would caution against the conclusion that the overall manufacturing sector has bottomed until both PMIs are clearly rising. Over the past month, the most notable development in China’s equity market has been the near-vertical outperformance of A-shares versus the global benchmark. A catch-up period for A-shares was arguably warranted given the sustained rally in investable stocks since early-November, but Chart 2 highlights that the speed of the recent rise has pushed relative A-share performance quickly into overbought territory. At a minimum, a period of consolidation over the coming few weeks is likely. Chart 2Too Far, Too Fast​​​​​​​ The relative performance of EM stocks ex-China is one of the equity components of our BCA Market-Based China Growth Indicator, which has recovered over the past few months. However, Chart 3 highlights that the performance of EM ex-China reliably led Chinese investable stocks since the beginning of last year, and are now raising a red flag. A near-term relapse in investable equity performance would be consistent with our view that earnings face further downside risk over the coming few months. ​​​​​​​Chart 3EM Ex-China Is Flashing A Warning Sign For Chinese Investable Stocks Within the investable equity market, our low-volatility sector portfolio remains in an uptrend versus the broad market, although the composition of this portfolio has shifted significantly over the past few weeks. Financials, industrials, and energy stocks now account for 86% of our long MSCI China Low-Beta Sectors / short MSCI China trade, which is likely surprising to many investors given their traditionally cyclical characteristics. Chart 4 highlights that the relative performance of our low-beta trade has exhibited a reliably counter-cyclical message; this, in combination with the fact that it remains above its 200-day moving average, signals that it is still premature to shift to a cyclical overweight stance favoring Chinese stocks. Chart 4No Green Light Yet From Low-Vol Stocks Value stocks have been responsible for more of the rally in China’s investable market versus the global average than their growth peers (Chart 5). This underscores that at least part of the rise in investable performance has been due to a relative valuation trade, rather than strong conviction that the Chinese economy will strengthen materially over the coming year. Chart 5The Rally Has Been Led By Cheap Stocks Table 2 highlights that the 3-month interbank repo rate is down materially from its 12-month high, a decline that is now passing through into lower bank lending rates. According to the PBOC, the weighted average lending rate declined 30 basis points in Q4, after having been essentially unchanged in Q3. The decline validates our model for predicting the rate, which had been calling for a non-trivial decline. Despite the continual expression of concern in the financial press about rising onshore corporate bond defaults, spreads on SOE corporate bonds have been steady over the past 6 months. Spreads remain elevated when compared with late-2016 levels, but the recent trend in spreads does not suggest that domestic financial conditions are getting tighter. Chart 6 shows that the recent rise in CNY-USD is consistent with a tariff-based framework that we had presented for the exchange rate several times last year. While the rate was on its way to breaking through the psychologically important level of 7 for USD-CNY, trade talks with the U.S. have helped the rate rise to a point that is consistent with the current tariff regime. CNY-USD has already overshot to the upside based on interest rate differentials, but Chart 6 implies that further gains may occur if tariff rollbacks are part of an eventual deal with the U.S. Chart 6CNY-USD May Rise Materially Further If Tariffs Are Rolled Back Jonathan LaBerge, CFA, Vice President Special Reports jonathanl@bcaresearch.com Cyclical Investment Stance Equity Sector Recommendations

Highlights Analysis on Indonesia is available below. EM financial markets have diverged from the global growth indicators they have historically correlated with. This raises doubts about the sustainability of this rally. In China, broad bank credit has not accelerated at all, while non-bank credit growth rose sharply in January. The lack of recovery in broad bank credit growth is corroborated by lingering sluggishness in broad money growth. This refutes widespread perception in the global investment community that Chinese banks have re-opened the credit spigots again. Feature The headline news has all been positive for emerging markets over the past two months: The Federal Reserve is going on hold, China is stimulating its economy, the U.S. and China are nearing a trade agreement and risk-on market dynamics are permeating worldwide. Nevertheless, EM stocks have failed to outperform the global equity benchmark (Chart I-1, top panel). Notably, EM relative equity performance rolled over in late December when global share prices bottomed. Chart I-1EM Stocks Have Underperformed DM Ones Since Late December In absolute terms, EM equities have been attempting to break above their 200-day moving average, but have so far failed to do so decisively (Chart I-1, bottom panel). When a market struggles to break out or outperform amid favorable news flows and buoyant investor sentiment, the odds are that it is facing formidable headwinds under the surface, and is at risk of relapsing. We sense EM currently fits this profile. Needless to say, investor consensus is very bullish on EM, and dominated by the above-mentioned narrative, specifically the Fed turning dovish and China stimulating, which is reminiscent of 2016 when EM staged a cyclical rally. Consequently, investors have rushed to pile into EM stocks and fixed-income. Chart I-2 illustrates that asset managers’ net holdings of EM ETF (EEM) futures have doubled since October 2018. Chart I-2Investor Consensus Is Very Bullish On EM As of mid-February, EMs were by far the most overweight region within global equity portfolios, according to the most recent Bank of America/Merrill Lynch survey. The survey states that net 37% of global equity investors - who participated in the survey - were overweight EM. One of our clients that we met with on the road last week summed it up like this: “Investors have ‘recency bias’.” In other words, investors believe that 2019 will resemble 2016, and in turn have no appetite to bet against Chinese stimulus. We are in accord with this interpretation of investor behavior and the EM/China rally. Yet there are some noteworthy differences between today and 2016. First, in 2016, there was massive stimulus for China’s property market. At the time, the People’s Bank of China (PBoC) monetized the unsold housing stock in Tier-3 and -4 cities via its Pledged Supplementary Lending facility. At present, there is no stimulus for real estate. Second, by early 2016 EM profits had already contracted substantially. EM profits have yet to shrink in the current downtrend. Our thesis is that EM profits will contract this year for reasons we elaborated on in depth in our previous report, Mind The Time Gap. China’s credit and fiscal impulse leads EM/Chinese profits by about 12 months, and the recent improvement in this indicator, if sustained, suggests that a trough in EM/Chinese corporate earnings will only be reached in late 2019 (Chart I-3). Therefore, as EM profits shrink, investors will likely sell EM risk assets. Chart I-3EM Corporate Earnings Are Beginning To Contract Altogether, these differences with 2016 make us reluctant to chase the current EM rally, and we continue to expect a meaningful reversal in EM risk assets in the months ahead. Monitoring Global Growth We maintain that EM is much more leveraged to global trade and China’s growth than to Fed policy. For a detailed discussion on this matter, please refer to EM: A Replay of 2016 or 2001? report from February 7, 2019. Therefore, the Fed’s dovish turn is not a sufficient reason to buy EM risk assets. To buy EM cyclically, we would need to change our outlook on global trade and Chinese imports. China influences the rest of the world via its imports. A closer look at the indicators that correlate with EM risk assets and commodities do not justify the recent EM rebound. In particular: The import sub-component of China’s NBS manufacturing PMI strongly correlates with EM share prices, excess returns in EM sovereign credit, and industrial metals prices and suggest that investors should fade this rebound (Chart I-4). Chart I-4EM Stocks, EM Credit Markets, As Well As Commodities Prices Are Driven By Chinese Imports The Caixin manufacturing PMI for China was up in February, but the NBS manufacturing PMI fell. In turn, manufacturing PMI indexes in Korea, Taiwan, Japan and Singapore are all plunging, with several of them dropping well below the 50 boom-bust mark (Chart I-5). Chart I-5Asian Manufacturing Is Contracting Korean, Taiwanese, Japanese and Singaporean shipments to China were shrinking in January, while their exports to the U.S. were resilient (Chart I-6). This confirms that global trade has been weak due to China, and that there are no signs of its reversal. Chart I-6Asian Exports To China And U.S. Moreover, Korea released its February export data, and its aggregate outbound shipments are contracting (Chart I-7). Chart I-7Korean Exports: Deepening Contraction Further, China’s container freight index – the price to ship containers – has rolled over again after picking-up late last year due to front-loading of shipments to the U.S. which were induced by the U.S. import tariffs. This signals ongoing weakness in global demand, and does not justify the latest rebound in EM financial markets in general and currencies in particular (Chart I-8). Chart I-8Global Trade Is A Risk To EM Currencies Finally, even in the U.S. where manufacturing has been the most resilient globally, the odds point to notable weakness in this sector. Specifically, the continuous underperformance of U.S. high-beta industrial stocks to U.S. overall industrials beckons a further slowdown in American manufacturing (Chart I-9). Chart I-9U.S. Manufacturing Is In A Soft Spot Bottom Line: Although financial markets are forward-looking, the recent rally has been too fast and has already gone too far. This has created conditions for a material setback as global/China growth will continue to disappoint in the months ahead.  China: Credit Versus Money Growth We have been receiving questions from clients as to whether investors should heed to the message from China’s money or credit data, given they are presently sending contradictory messages (Chart I-10). Chart I-10China: Narrow, Broad Money, And Aggregate Credit Even though narrow money (M1) has historically been an excellent indicator for China/EM business cycles, the most recent (January) print – M1 annual growth rate registered a record low – was distorted due to technical/seasonal factors, and should be ignored. Specifically, deposits by enterprises plunged in January and household deposits surged as companies paid out bonuses to employees in late January ahead of the Chinese New Year that began on February 5 (Chart I-11). Provided enterprise demand deposits are in M1 but household demand deposits are a part of M2, M1 was artificially depressed in January. It will rebound in February. Chart I-11China: Technical Reasons For M1 Plunge In January Broad money provides a more comprehensive picture of money creation in China. As such, it is more relevant to compare broad money with aggregate credit. To compute aggregate credit, we add outstanding central and local government bonds to Total Social Financing (TSF). Chart I-12 illustrates the latest improvement in aggregate credit is not confirmed by either the PBoC’s broad money measure, M2, or our measure, M3 (M3 = M2 plus other deposits plus banks’ other liabilities excluding bonds). We created this M3 measure of broad money supply because in our opinion, M2 has been underestimating the extent of money creation in China in recent years due to financial engineering. Chart I-12The Recent Uptick In Aggregate Credit Is Not Confirmed By Broad Money As discussed in Box I-1 on pages 12-13, lending or purchasing of securities by banks simultaneously creates money. Therefore, bank broad credit acceleration should be mirrored in a broad money upturn. Does the lack of revival in broad money mean the latest uptick in aggregate credit data has been driven by non-bank credit? Our analysis suggests yes – non-bank credit is responsible for the strong rise in the aggregate credit numbers in January. We deconstructed aggregate credit into broad bank credit and non-bank credit (Diagram I-1). Chart I-13 illustrates that broad bank credit has not accelerated at all, while non-bank credit growth rose in January. Chart I-13China: Recent Credit Acceleration Is Due To Non-Bank Credit The lack of recovery in broad bank credit growth is corroborated by lingering sluggishness in broad money (both M2 and M3) growth (Chart I-14). Chart I-14Broad Bank Credit Is Consistent With Broad Money (As It Should Be) Consequently, this refutes the widespread perception in the global investment community that Chinese banks have re-opened the credit spigots. Chart I-15demonstrates the annual growth rate of each component of broad bank credit. While mainland banks’ loan growth to enterprises has accelerated, their lending to non-bank financial institutions has continued to shrink.  Chart I-15Broad Bank Credit And Its Components In sum, broad bank credit and broad money have not revived, and their impulses are rolling over, having failed to break above zero (Chart I-14, bottom panel). Bottom Line: The improvement in aggregate credit growth in January was due to credit provided/bonds purchased by non-banks rather than by banks. This does not tell us whether the credit growth acceleration is sustainable. For a more detailed discussion on the differences between money and credit, please refer to Box I-1 on page 12-13. Investors prefer simple narratives, and have readily embraced the story that China has opened up the credit faucets. Broad bank credit data and broad money supply data do not corroborate this thesis. It may change in the months ahead, but our point is that for the moment there is not yet a simple narrative about China’s credit cycle. Investment Implications Even though China’s aggregate credit impulse ticked up in January, the 2011-‘12 and 2015-‘16 episodes signify that its bottoming can last many months. Critically, EM financial markets have historically lagged turning points in the aggregate credit impulse. These time lags have been anywhere between three to 18 months over the past 10 years. Furthermore, in 2012 there was only a minor rebound in EM share prices – not a cyclical rally – in response to the significant rise in China’s aggregate credit impulse (Chart I-16, top panel). Chart I-16Beware Of The Time Lag Hence, even if January marked the bottom in the aggregate credit impulse – which is plausible in our opinion – EM risk assets will remain at risk based on historical time lags between the aggregate credit impulse and China-related financial markets.1 BOX 1 Why And When Money Supply Differs From Credit The following elaborates on the key differences between broad money supply and aggregate credit.  1. Why and when do broad money and credit diverge?  When commercial banks provide loans to or buy bonds (or any other asset) from non-banks, they simultaneously create new money supply/deposits. Broad money supply is the sum of all deposits in the banking system, which is why we use the terms money and deposits interchangeably. When non-bank financial institutions – in China's case financial trust and investment corporations, financial leasing companies, auto-financing companies and loan companies – as well as enterprises and households make loans or buy bonds, they do not create money. Hence, money supply/deposits is mostly equal to net cumulative broad bank credit creation. The difference between aggregate credit and money supply is due to lending activities of non-bank entities (see Diagram I-1 on page 9). Lending, purchasing of bonds, or any other forms of financing by non-bank entities does not change money supply. Thus, aggregate credit is more relevant than money supply to forecast business cycle fluctuations. Apart from the fact that banks still play a very large role in aggregate financing in China, there are a few other reasons why one should not ignore broad money and rely solely on aggregate credit: Banks can extend credit, but might choose not to classify it as loans on their balance sheet for regulatory reasons. Chinese banks did this in the past by booking loans as non-standard credit assets. In any case, when a bank lends to a non-bank it creates new deposits/money, and it is hard to conceal deposits/liabilities. In these cases, broad money supply gives a better signal about the true extent of credit growth than statistics on loans. If under regulatory pressures banks reclassify their non-standard credit assets as loans, the amount of loans will expand, even though no new lending occurs. Yet, money supply/deposits will not change. In this case, loan numbers will give a false signal and money supply will be a better indicator for new credit origination by banks and, thereby, for economic activity. The true measure of Chinese bank loans and credit data were probably disguised over the past several years because banks and non-bank financial institutions were involved in financial engineering. However, in the past two years, the regulatory clampdown forced Chinese commercial banks to unwind some of these structures and properly reclassify items on their balance sheets. Both the masking of credit assets and the ensuing reclassification could have distorted loan and credit data. This is why we use broad money supply as a litmus test to gauge banks’ broad credit origination. Given TSF includes bank loans but does not include banks’ non-standard credit assets, we believe TSF understates the amount of credit in the economy. As a result, we have not been able to calculate an accurate aggregate level of non-bank credit. Only since mid-2017, when under the regulatory clampdown, banks have stopped classifying loans as non-standard credit assets, can the annual growth rate of TSF serve as a meaningful statistic. Hence, we estimate the annual growth rate of non-bank credit only starting in 2018 (please refer to Chart I-13 on page 9). 2. Does the central bank (PBoC) create money by injecting liquidity into the system? Barring lending to or buying assets from non-banks – which does not typically occur outside of quantitative easing (QE) programs – central banks do not create broad money or deposits. Central banks create banking system reserves, which are not part of the broad money supply in any country. Money supply/deposits, the ultimate purchasing power for economic agents, is created solely by commercial banks “out of thin air,” as we have discussed and illustrated in our series of reports on money, credit and savings. 3. Why do we use impulses (second derivatives of money/credit) rather than growth rates? Our goal is to forecast a change in economic activity/capital spending/imports/enterprise revenues – i.e., a change in flow variables. Money and credit are stock variables. Therefore, a change (the first derivative) in outstanding money and credit produces flow variables. The latter measures new credit and money origination in a given period. These are comparable with flow variables like spending, income and profits. To gauge changes in flow variables, i.e., the growth rate of spending, one needs to calculate a change in new money and credit origination – i.e., change in their net flow. In brief, to do an apples-to-apples comparison, one needs to use the second derivative (a change in change) in money and credit – i.e., changes in their flows – to predict changes in flow variables such as GDP/capital spending/imports/enterprise revenues.   Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com Lin Xiang, Research Analyst linx@bcaresearch.com   Indonesia: It Is Not All About The Fed Indonesian stocks have outperformed their emerging market peers significantly in the past few months as the Federal Reserve has turned dovish and U.S. rate expectations have declined. Although U.S. bond yields do strongly and inversely correlate with Indonesian stocks’ relative performance versus the EM equity benchmark (Chart II-1, top panel), we believe there are other factors – such as Chinese growth and commodities prices – that are also important to this market (Chart II-1, bottom panel). Chart II-1Indonesian Stocks: The Fed Versus Commodities In the next several months, slowing Chinese growth, lower commodities prices, and a renewed sell-off in EM markets will take a toll on Indonesian financial markets. Indonesian exports are contracting which will intensify as commodities prices fall and China’s purchases of coal and base metals drop (Chart II-2, top panel). Chart II-2Indonesia: Exports Are Shrinking Indonesia’s current account deficit is already large and will continue widening as the export contraction deepens (Chart II-2, bottom panel). Remarkably, the nation’s commercial banks have been encouraged to keep the credit taps open as the central bank – Bank Indonesia (BI) – has been injecting enormous amounts of liquidity (excess reserves) into the banking system (Chart II-3, top panel). Given these liquidity injections, bank credit and domestic demand growth have remained more resilient than would otherwise have been the case. Chart II-3The Central Bank Is Injecting Liquidity Yet, by injecting such enormous amounts of excess reserves into the system, the central bank has more than negated its previous liquidity tightening, resulting from the sales of its foreign exchange reserves in order to defend the rupiah (Chart II-3, bottom panel). The implications of such policy are that these excess reserves could encourage speculation against the rupiah, especially amid weakening global growth and falling commodities prices. Provided foreigners own large portions of Indonesian stocks and local-currency government bonds, a depreciation in the rupiah will produce a renewed selloff in the nation’s financial markets. A final point on Indonesian commercial banks: their net interest margins have been narrowing sharply (Chart II-4, top panel). Chart II-4Commercial Banks' Profits Will Weaken Moreover, as global growth slows, non-performing loans (NPLs) on the balance sheets of Indonesian banks will rise. In turn, provisioning for bad loans will also increase, and bank earnings will decline (Chart II-4, bottom panel). These dynamics will be bearish for Indonesian commercial banks, which account for 44% of the overall MSCI Indonesia index. Bottom Line: Continue avoiding/underweighting Indonesian stocks and fixed-income markets. We continue shorting the IDR versus the U.S. dollar. Ayman Kawtharani, Associate Editor ayman@bcaresearch.com Footnotes 1      Please note that this represents the Emerging Markets Strategy team’s view and is different from BCA’s house view on global risk assets and global growth. The key point of contention is the outlook for China’s growth.   Equity Recommendations Fixed-Income, Credit And Currency Recommendations